Confirmatory cytopathology and potential impact on the predictive value of GEP

Abstract

Purpose : To compare rates of metastasis in patients with uveal melanoma (UM) who underwent gene expression profiling (GEP) [DecisionDx-UM, Castle Biosciences] and confirmatory cytopathology (CP) to patients who had GEP alone.

Methods : Patients diagnosed with UM at Mass Eye and Ear who underwent fine needle aspiration biopsy prior to proton therapy between 2013-2019, were included. Patients were divided by CP status: those with CP results confirming or suspicious for UM along with GEP (GEP+CP group), and those who had no CP or non-diagnostic CP results (GEP group).

Results : One-hundred-eight patients had GEP results, 74 (68.5%) with CP and 34 (31.5%) without CP. Patients in the GEP+CP group were older (median age: 60.0 v. 50.2 years, P=0.02) with larger tumors (median tumor thickness 5.3 v. 3.0 mm, P=0.0005; largest basal diameter 14 v. 13.8 mm, P=0.91) than those in the GEP group. Median follow-up was 3.6 v. 5.0 years in the GEP+CP and GEP groups, respectively (P=0.07). In the GEP+CP group, 19 (25.7%) patients were diagnosed with metastasis versus 7 (20.6%) in the GEP group (P=0.38). GEP class for the GEP+CP and GEP groups, respectively, were as follows: 1A: 29 (39.2%) v. 14 (41.2%); 1B: 16 (21.6%) v. 12 (35.3%); 2: 29 (39.2%) v. 8 (23.5%) [P=0.21]. Median time from UM diagnosis to metastasis in the GEP+CP and GEP groups, respectively, was 21.3 v. 24.5 months in patients with class 1A tumors (P=0.80), 68.6 v. 37.8 months in class 1B tumors (P=1), and 21.1 v. 19.6 months in class 2 tumors (P=0.82). Metastasis was rare in class 1A and 1B tumors. In those with class 1A tumors, 3 v. 2 patients developed metastasis in the GEP+CP and GEP groups (10.3% v. 14.3%, P=0.53); and 1 patient developed metastasis in each group (6.3% v. 8.3%, P=0.68) among class 1B tumors. Among patients with class 2 tumors, 15 (51.7%) developed metastasis in the GEP+CP group compared to 4 (50.0%) in the GEP group (P=0.62). The 5-year rate of metastasis in patients with class 2 tumors was 54.9% (95% CI: 34.5-77.6) in the GEP+CP group and 50.0% (95% CI: 22.5-84.8) in the GEP group (P=0.97).

Conclusions : Within each GEP tumor class, we did not find significant differences in rates of metastasis between patients with and without confirmatory CP. Such findings suggest GEP is adequate for risk prediction. Additional analyses of a larger cohort to validate these findings will be beneficial.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.